Compared
to the stock circuit in Fig. 1, the EQ circuit depresses the output in the 1.8kHz to 5kHz region slightly, and then lifts
the remainder of the HF output several dB. The input impedance of the circuit was unchanged.
Parallel Possibilities
I
then moved on from series topology and tried several promising parallel configuration models with CALSOD. They were promising
in the sense that they could result in a filter with the desired driver acoustic center phase correction at crossover, and
hopefully an improved overall sound. I was specifically interested in 3rd order high/low pass Butterworth and 4th
order high/low pass Butterworth filters for this driver.
A
strange thing happened though - when I modeled each parallel filter its input impedance rose dramatically at an intermediate
frequency. For the 3rd order BW the impedance rose to 42 ohms at 650Hz; for the 4th order BW it was
45 ohms at 500Hz. The only way to approach the input impedance of the series filter was to add more components to these parallel
filters in the form of conjugate networks for control of the woofer’s impedance and to counter the rising inductance
of the compression driver’s voice coil. The conjugate networks, when added, were successful in restoring reasonable
input impedance to the filters, but now I had networks with three to four times as many components in them as the N-1500A,
and yet had very similar performance.
Bottom Line
The
N-1500A filter is hard to equal, much less beat, when using competing parallel type designs with the 604E. An N-1500A filter
with upgraded components and an MF EQ circuit added is probably the most cost effective crossover improvement one can make
for this classic duplex.
Getting the magnets re-charged on 42 year old alnico drivers really helps too!
Jeff Markwart 4/26/2008
